[Wafer: response was truncated before the model finished its internal reasoning. Increase max_tokens, or disable thinking on this model (e.g. chat_template_kwargs.enable_thinking=false), then retry.]

Wall Street's once-reliable investment formula - "just buy the M7" - is breaking down. Meta and Tesla are down this year, while chip and server makers that supply the AI buildout are posting triple-digit gains. That divergence has investors hunting for a new set of market leaders.

The M7 - Nvidia, Apple, Microsoft, Alphabet, Amazon, Meta, and Tesla - posted average annual share-price gains of 111.3% in 2023 and 60.2% in 2024. This year, performance has sharply diverged. According to the New York Stock Exchange, Meta is down 10.5% from the end of last year and Tesla is down 28.9%. Even Amazon, the top M7 gainer, is up just 17.9%.

By contrast, AI infrastructure companies have climbed steeply. SanDisk is up 430.2%, Dell 247.7%, and Micron 208.8% over the same period.

The P7: AI's suppliers take center stage

That performance gap has put a new grouping in the spotlight: the "P7," short for "Parabolic 7," proposed in June by Ben Emons, chief investment officer of FedWatch Advisors. The P7 comprises SanDisk, Marvell, Micron, Intel, Dell, AMD, and Broadcom.

These companies supply memory, AI chips, custom silicon, network equipment, and servers. Their revenue grows as Big Tech spends more on data centers and AI infrastructure. If the M7 are the buyers investing in AI, the P7 are the suppliers that convert that investment into earnings.

The shift is visible in market-capitalization rankings. Meta and Tesla have slipped to 10th and 11th globally, at $1.502 trillion and $1.261 trillion respectively, while Broadcom has grown to roughly $2 trillion. The Financial Times assessed the trend bluntly, saying the M7 are "no longer great."

Revenue beats spending

Behind the changing leadership is a shift in how investors evaluate AI. Large-scale AI investment plans used to be enough to earn growth recognition. Now, the market wants to see that spending converted into revenue and cash flow.

The difference showed in recent M7 earnings. Microsoft's 365 Copilot paid users grew to 30 million and Azure revenue rose 43%, sending the stock up more than 15% after its report. Amazon's AWS posted second-quarter revenue growth of 37% year-on-year, beating the market forecast of 31%; its AI business has reached an annualized scale of about $25 billion.

Alphabet, by contrast, saw its share price fall more than 7% after its earnings as profitability concerns surfaced despite strong cloud growth. Meta and Tesla continue heavy investment in AI, robotics, and autonomous driving, but the uncertain timing of monetization is weighing on their stocks.

As Big Tech's AI investment scale grows, free cash flow has become a key measure. The market appears to be asking not how much is invested in AI, but how much of that investment returns as profit.

Growth forecasts are also cooling. According to Reuters, the M7's quarterly year-on-year revenue growth is expected to fall from 35.1% in the third quarter to 23.1% in the fourth quarter, and to 7.1% in the first quarter of next year.

Wall Street hunts for the next leaders

Wall Street is actively seeking new groupings. Bank of America proposed an "AI Big 10" that adds AMD, Broadcom, and Micron to the existing M7. A name called "MANGOS" - combining Microsoft, Alphabet, and Nvidia with Anthropic, OpenAI, and SpaceX - has also surfaced.

The P7 is the candidate drawing the most recent attention. It remains uncertain who will ultimately win the AI services race, but as long as Big Tech keeps competing, investment in chips, memory, servers, and network equipment is likely to continue. Goldman Sachs recently assessed that the rise of AI-related stocks is based on growth in corporate earnings rather than mere expectations, naming AI infrastructure and power infrastructure as key investment themes for the second half.

Why this matters for finance professionals

For investors, the practical takeaway is to distinguish between AI spenders and AI earners. The M7's revenue growth is decelerating, while suppliers like the P7 are converting Big Tech's capital expenditure directly into income statements. That suggests a closer look at free cash flow trends and revenue conversion rates - not just AI spending totals - when evaluating both groups. The era of treating the M7 as a single, safe bet is over.